3 mobile app statistics and analytics platform comparison



This article by Lei Feng Network finishing from ZDNet, the content has changed, only for you to be enthusiastic about the mobile application development promotion Operation friend Reference.

At present, the market is more common, using a more mobile application statistics platform about 3, 4, foreign more popular is flurry, the function is very comprehensive; the other is Google Analytics also launched a mobile version, because you understand the reasons, in the domestic basic can not be used normally. But the domestic statistical analysis platform is now more famous is the Friend Alliance and TalkingData.

In addition, some application markets also have specialized statistical analysis tools to provide data services to application developers on their developer platforms.

We will focus on flurry, friends Alliance, and TalkingData, the three mobile application data analytics platform to make a simple contrast experience.

Flurry

As a benchmarking platform in the field of mobile application statistics, flurry has a very comprehensive range of functions, not only for data statistics and analysis functions, but also for app Circle advertising and promotion platform functions. It can be said that flurry is currently the most comprehensive mobile application statistical analysis products, in addition to statistics in a single application of various data indicators, but also to provide cross-application conversion statistics, such as for enterprise-level user capabilities. Flurry is also in a leading position, simply from the data statistics function of mobile applications. Its function module is set up reasonably, the analysis dimension is comprehensive, the analysis flow is easy to understand, is the strongest.

Figure 1:www.flurry.com

In the case of single-application data statistical analysis, flurry classifies various analysis modules to make it easier to understand the data. After entering the application analysis, the user first sees is the dashboards panel, the content includes some basic trend data about the application, such as the conversation quantity trend (many parameters of flurry use "session"-sessions this concept), activates the user number change trend, Basic data such as average session duration. Dashboard users can quickly understand the current overall trend state of their apps.

Recently flurry to dashboard has been strengthened, now flurry users can have their own needs to create dashboard panels, flexibility has greatly improved. And Flurry provides a number of ready-made dashboard templates, corresponding to several more common types of applications, users can quickly select their desired dashboard style, easy to monitor the desired data.

Figure 2:flurry's dashboard function is very powerful

In most cases, dashboard is basic enough for the data reporting requirements. But to further understand the user, understand the channel and other content, only by the simple dashboard of data is not enough. Flurry is very comprehensive in this area, with a total of 4 categories and 23 features to help developers understand users and understand user behavior. 4 large categories include: usage, audience (audience), events, and Technology (Technical). The main statistics in "use" are the basic conditions of some applications, including new users, active users, sessions, session duration, frequency of use, and so on.

It should be explained that the duration of the session, user frequency, user retention three projects, if the developer provided the corresponding system platform market developer ID, you can compare their own application and the same classification under the average level of comparison in these three projects, Gives developers an overview of where and how their applications are located in the industry. One of the audience's interests, such as audience interest, is divided into two categories, one is the current application of the user's interest in other developers, and the other is the current application of the user's interest in the same type of application, also includes the user's age, gender, geographical location, language information, This information, in addition to allowing developers to understand their own app's audience attributes, can also be used as a unique user group Flurry feature.

Figure 3: Only registered users can get full functionality

Event classification is primarily a developer-defined trigger event, known to the industry as a "custom event," which is used primarily to monitor user-specific actions that some developers care about, such as whether to pay or not, but also the key to monitoring the various conversions within the application. This is the statistical platform is flexible enough to directly reflect, this aspect of flurry do quite well, not only can easily organize, add a variety of custom events, and can also be based on the trigger of custom events, give a separate visual icon, very intuitive.

With the flurry update, the funnel feature has been improved in the event, and developers can organize a series of custom events into a series of funnels to examine the user conversion rate between each step of the custom event through a custom event funnel. This is a very flexible statistical analysis method, especially suitable for the IAP in the application of monitoring in-app items sales and promotion situation. The technical classification of the operator information and equipment used by users, such as networking, system version, error reporting, and so on, mainly with the user group division. Finally, Flurry provides a complete management module, developers can define their own user groups in the Management module, custom events, the latest update is added to the early warning function, by setting the warning threshold and notification method, developers can know the unexpected situation anytime and anywhere, to respond immediately.

The basic function, flurry is very comprehensive, but also in the details of a lot of commendable points. such as fine User Group division, and convenient global multi-dimensional screening. Developers through the user Group Division module, according to their own needs to set up a different user groups, the process is actually through some conditions to filter out the user. such as the use of time, sessions and other information to determine the loyal users, through the specific billing event to identify paid users, or according to audience information, such as age, gender, the user group segmentation. When the developer completes the segmentation of the user base, it can easily lock their target users, conduct continuous tracking, analyze the behavior of the target user, and then carry out operational actions, or refine the product accurately.

The multi-dimensional filtering function provides developers with a powerful filtering tool, which allows developers to quickly get the data reports they need, and to understand their applications in a comprehensive and detailed way through a combination of user groups, time, and version information.

Flurry in both functional and detailed performance, in addition to the previously mentioned single in-app data analysis, the company page can also be found in the developer's current release of all the application data situation, the latest update added to the conversions (conversion) function, It also provides a variety of users, session conversions, and provides conversion funnel capabilities for developers, especially for developers and manufacturers of multiple applications. In addition, Flurry Appcircle also provides developers with application promotion services, developers can be used as a carrier for advertising, but also to join the promotion channel, the former to help developers profit, the latter is used to help developers to promote the application, can be said to consider very comprehensive.

Figure 4: Complete cross-application conversion support is one of the most powerful features of flurry

For domestic developers, flurry has some innate deficiencies to be aware of.

First of all, the service platform is set up abroad, based on some well-known issues, in the domestic visit flurry is not stable, although much better than Google Analytics, access speed is acceptable, but this is only the front desk service, data statistics is the real core of data collection in the background, The front desk service is accessible and does not indicate that background data collection will be smooth. If you develop an application that focuses on overseas markets, users are basically European and American users, then the accuracy of flurry data can be guaranteed; Conversely, if the target market is domestic, then the problem of data accuracy and data delay should be considered.

Flurry does not provide a Chinese interface like Google Analytics, it only has an English UI, and even if the English language is good, it takes a lot of time to understand how many of the concepts and parameters are going around.

Figure 5:flurry's appcircle features help developers promote their products

Friends League

Friends Alliance is one of the projects of Innovation Factory incubation, and it is the most familiar mobile application data statistic and analysis platform for domestic developers. At some level, there are many similarities between friends and flurry. But in terms of comprehensiveness and completeness, there is still a certain gap. There are a number of SDKs available, including the standard statistical analysis SDK, the SNS Sharing SDK, the user Feedback SDK, and the Application Alliance SDK. Developers can choose their own choice, but the statistical analysis SDK is a must-have option. On the platform side, AU currently supports iOS, Android and Windows Phone platforms (currently the Windows Phone platform only provides the statistical analysis SDK).

Figure 6:www.umeng.com

Friends of the league also made a major version update, with a redesign of UI and feature partitioning. Added some new features, such as the original "Return users" section was replaced by the current retained users, and added the original channel analysis module is not. In addition, the Friends of the league's update has greatly adjusted the UI design, the Platform homepage and feature pages are all updated. For the old users, such a major change, in use will inevitably encounter the problem of re-adaptation. Have to say is, friends of the Alliance after the revision, in the function module settings more like flurry.

Figure 7: Friends League Product Center

The changes in the UI are not really the focus, and it is more practical for developers to provide perfect data statistics. After the revision of the league did not provide a "company" page, like Flurry, but still provide a simple comprehensive product trend chart, enter the product center, you can see all the product concise data profile, and based on new users, active users, and the number of start times to provide 3 selectable time granularity of the graph. After clicking on a specific application, you will be presented with a single Application data analysis page.

In the case of statistical analysis of single application data, the functions provided by friends Union are quite comprehensive. Its module design thinking is also relatively clear, basically along the basic situation, user situation, equipment situation, event monitoring, transformation analysis of the process continues to deepen. Statistical profiles include basic statistics and version distributions to help developers understand the basic data situation, mainly information such as user growth and the number of application launches. In subsequent user analysis, AU has added two new parameters to active users and retained users in the latest update.

In active user charts, developers can learn about the daily active, weekly, and monthly active users of the app, as well as the weekly activity rate and monthly activity rate. And the retained user is used to investigate the user loyalty parameters, simply speaking, the new user after a period of time, still have how many users continue to use the app. This is a very important parameter, on the one hand can reflect the quality of the application, on the other hand, it is also an important parameter to investigate operational action, channel promotion and channel quality. For example, a developer can evaluate the operational details of the activity promotion effect, channel user quality, etc. by examining the retained user situation at a certain point in time, and comparing the new user situation gained by application promotion.

In terms of time granularity, the Friend Alliance's retained user module provides two ways of weekly and monthly queries. But for many applications that are now iterative in weeks, this granularity is not fine enough. After all, a large number of customer churn occurs in the first few days after the installation of the application, it is necessary to see the retention rate of the day.

Another important feature of user analysis is page access statistics, and developers who have used flurry will not be unfamiliar with this feature. This function is mainly statistical users in the application of the page jumps, can be more convenient to understand what features, pages are more popular with users, but also easier to find a serious user Churn function module, the product improvement has great value.

Another feature added in this update is the channel analysis module. This is mainly for the domestic third-party Android app promotion channel more, developers want to understand the different channels bring users, and the emergence of the function (flurry there is no channel analysis function). Friends of the channel analysis to provide users with different channels of new, active, start the number of basic parameters such as statistical functions, combined with the previous user retention module, the developer can be relatively intuitive to understand the quality of the promotion channels. It is worth mentioning that the alliance is dedicated to developers to provide "channel packaging tools", so that developers for different channels to package applications, to simplify developer development work.

Figure 8: Friends Alliance provides a complete channel packaging tool for Developers

The Friend Alliance custom event module is primarily used to manage set up custom events and provides data statistics for each custom event. If developers need to understand the conversion between events, they need to enter the conversion rate Analysis module, using the funnel model provided by friends, the basic use process and flurry very close, set up the event conversion process, you can see the conversion between each step.

Figure 9: Convenient conversion funnel settings

These are some of the key features recently updated by Friends Alliance, as well as the traditional terminal and network, error analysis and other functions, Friends of the Alliance is still available to developers. These data can not be used to subdivide the user group like Flurry, hope that friends can launch the user group segmentation function as soon as possible, on the one hand can provide more comprehensive services to developers, on the other hand, but also more fully utilize their own platform data, make these data become organic whole.

Friends League in the latest update, added a similar to the flurry of the global filtering function, but now this function seems to have some problems, such as choose to view by version, and then select a time period, the previously selected version settings will revert to the default "All" and vice versa. If this is not a bug, then it can only be this global filtering function and can not be cross-version, cross-time period cross-filtering, but rather like the global filter switch.

Friends of the league in the statistical function of the set, or relatively comprehensive, but at the data analysis level, and flurry there is still a large gap, there is still a lot of space to dig.

As a high-profile mobile Application data statistical analysis platform, friends of the league has done very good. In particular, the League of Friends and Flurry, with the application of the Alliance platform, on this platform users can be built-in friends of the mobile advertising platform to achieve profitability, but also through the Friends of the Alliance's Exchange network to achieve the application of promotion, for developers is still more convenient.

But domestic application promotion, advertising field can be said to be a mixed bag, the situation is quite complex. Although the service is convenient for developers, the value of the final implementation level needs to be further explored.

Figure 10: Affiliate APP Alliance supports AD delivery and volume promotion

Talking Data

This is a recently emerging mobile application statistical analysis platform, online less than 2 months, is still in the invitation testing phase. As a "new thing" in the field of mobile application statistics, talking data has its own unique statistical analysis system. For example, we mentioned a new feature recently updated by the Friends league--retention user and channel analysis, talking data has provided complete support. Although there are shortcomings in the channel analysis function, it is not difficult to see talking data in the product design has a unique perspective.

Figure 11:www.talkingdata.net

Talking data statistical analysis module has only 4 large categories, namely user and usage, participation analysis, channel statistics and custom events.

All the statistical analysis modules of talking data are housed in these 4 categories, which differ greatly from other statistical analysis platforms. And these 4 categories happen to be the same, from basic data statistics to deep data analysis such a process. This should be a product design created by talking Data's understanding of mobile application Analytics.

After entering the single Application statistical analysis page, the first thing to see is the application overview, this module is actually a quick report page, which embodies the user's attention to some of the indicators and parameters, which is close to flurry dashboard, but talking Data does not provide custom module functionality. Currently, we can learn about recent profiles, application summaries, active profiles, time-of-day analysis, and 30th data trends from the application overview, plus 4 top 10 ranking tables, including region, channel, launch count, and full version of the feature page. The data provided on this app profile page is sufficient if it is used only for daily data reporting.

In addition, the above mentioned charts have different classification options, such as in the 30th Data Trend chart, the developer can use the tag button to the new user, active users and other parameters switch between, while the time analysis also provides a comparison function, you can filter the items to see, and then check the target, To quickly see data differences between different dates in the same segment.

Figure 12:talking Data's quick report is very comprehensive

If you need a deeper understanding of the application, it is not enough to rely on the application profile. Through the user and the use of modules, developers can understand user trends, regional distribution, equipment models, operators and networks, error reporting, functional or relatively comprehensive. At the same time, all the charts in this module have different detail data to filter, such as trend analysis we can see the new users, active users, the average length of use and other information, and can be based on the date of data filtering.

The data in a detailed report of talking, such as the distribution of a device, provides both specific data and percentages, compared to the percentage provided by the League of Friends.

Figure 13: Trend analysis features that can be filtered by time period

Participate in the analysis, then provide user retention, user visits, the number of days to start, the length of Use, page access module. From the setup of these functional modules it is not difficult to see that the developer has entered the field of data analysis, these modules combined with the application of the developer operation, promotion, can analyze a lot of useful data. As we mentioned earlier, user retention is a feature that talking data has provided when it was launched 2 months ago, which is a good indication of its high level of attention. In the user retention module, there are two charts, namely user retention (i.e., retained users) and user churn funnel.

Talking Data's user retention provides a more granular analysis granularity, and developers can choose between day retention, weekly retention, and monthly retention as needed. More granular user retention features are more meaningful for application operations promotion-after a promotion operation, the next day you can see the new user's retention, the activity quality, channel quality inspection more valuable. At the same time, the new user churn funnel in days, more intuitive for developers to provide user churn, and this funnel chart can also be dragged and scaled by the mouse, this is really unexpected.

Figure 14: User retention features that support daily user retention statistics and user churn funnel help with granular operations

Talking data also has a page access module, but this part talking data performance is relatively general, although there are page jump statistics, but not enough intuitive table style for flurry and friends of the page jump stream, in the understanding and reading performance is not good enough.

Another feature that talking data pays more attention to is the channel statistics. Talking data is doing well in this respect. After adding channels, developers can filter the details based on the channel. After you expand the channel trend chart, you can view detailed data such as new users, active users, average hours of use, and so on, by using the data labels on the chart, which is extremely convenient for comparing channel quality. Combined with the previous first day user retention, is a tool to investigate the effectiveness of channel promotion. In addition to the channel data graph, the channel trend comparison icon provides a more intuitive comparison of new users, active users, and startup times, in the ointment that only the three items mentioned above can be compared.

Talking Data custom event module is very special, developers in the use of talking data custom events, do not need to advance the front-end platform for event declaration, just in the Integration SDK phase, according to the requirements of directly written custom events, talking Data automatically generates the corresponding custom event statistics project on the front-end service platform. Therefore, in the Talking Data custom event module, directly provides is an event conversion funnel, and this event conversion funnel creation is also very simple, as long as the drop-down list to select a good event, instant conversion data, easy and fast, and can be saved, easy to query in the future.

Figure 15: A custom event module that flows to clear and easy operation

Overall, talking data in the functional aspects of more emphasis on analysis, the overall guidance is in accordance with the "first statistics-after analysis-re-conversion" the process of progressive, can meet the needs of different levels of data statistics developers. Of course, the current talking data has many shortcomings, after all, the online time is relatively short. For example, talking data currently has a product center, but the product in the heart can only provide a simple cross-application cumulative number of users, the total number of new users today, such as very simple statistics, distance like the flurry provided by the kind of support across the application of conversion Statistics Product Center is still very far. The current talking data also does not provide user Group division function, as a focus on data analysis of a product, this need to strengthen.

In addition, the data screening is not uniform, although in the way of operation is relatively close, but also to provide a really effective screening, but the screening method is relatively single, which makes people feel more regrettable.

Summarize

More detailed parameter description, introduction, interested readers can go to these platforms on the product website to download the corresponding development documents, the development documentation has more detailed instructions and introduction.

There is also a famous Google Analytics has launched Mobile analytics, but Google's services are too unstable, open analysis sites are sometimes more difficult.

In fact, for a complex mobile application data statistics products, only look at the function is not enough, the stability of the service, data timeliness, and even technical support, etc., are the place to be considered.